Cloud circuit board detection data management system
By building a dynamic performance interaction model and nonlinear anomaly recognition module in the cloud circuit board detection system, combined with chaos theory analysis, the shortcomings of existing systems in identifying tiny abnormal signals in complex nonlinear dynamic systems are solved, and fault diagnosis is improved Accuracy and sensitivity prevent serious failures.
Patent Information
- Application Number
- CN202410968938.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing cloud circuit board detection systems perform poorly in identifying tiny anomalies in complex nonlinear dynamic systems, making it difficult to capture complex dynamic changes and nonlinear characteristics in circuit board detection data, resulting in potential failures not being discovered in time, which may lead to serious as a result of.
The cloud circuit board detection data management system is adopted, including signal acquisition module, dynamic model construction module, rate adjustment prediction module, nonlinear exception identification module, abnormal analysis module, optimization and adjustment feedback module and abnormal data archiving module, and dynamic circuit board construction Performance interaction model, combining chaos theory and stability analysis, identify and analyze abnormal signals, optimize monitoring strategies and parameter settings.
It significantly improves the system's ability to identify small abnormal signals, improves the accuracy and sensitivity of fault diagnosis, and promptly detects potential faults, greatly reduces the intervention time when the fault develops to a serious stage, effectively preventing possible serious consequences.
Smart Images

Figure FT_1 
Figure FT_2 
Figure FT_3
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of cloud data management, and in particular to a cloud circuit board detection data management system. Background Art
[0002] The field of cloud data management technology is a field that uses cloud computing technology to store, process and analyze data. In this field, data is uploaded to cloud servers and can be accessed from any location via the Internet. This not only improves the efficiency and flexibility of data processing, but also enhances data security and backup capabilities. Cloud data management technology supports a variety of applications, from simple file storage and sharing to complex database management and big data analysis, providing a wide range of services for enterprises and individuals.
[0003] The cloud-based PCB inspection data management system is a system that uses cloud-based data management technology to inspect circuit boards and manage inspection data. The purpose of the system is to ensure the quality and reliability of electronic products by efficiently and accurately detecting defects on circuit boards. By uploading inspection data to the cloud, it also allows remote access, analysis, and storage of inspection information, thereby achieving rapid fault diagnosis and quality control. This system aims to improve production efficiency, reduce errors in the production process, and ultimately achieve the effect of reducing costs and improving product quality.
[0004] Although existing cloud data management technology provides efficient and flexible data processing and storage capabilities for circuit board detection, and improves the speed of fault diagnosis and the controllability of the production process to a certain extent, there are still some shortcomings. Although traditional systems can identify and process abnormal data within a certain range, they are not ideal in identifying small abnormal signals in complex, nonlinear dynamic systems. Because traditional anomaly detection methods are often based on linear assumptions or simplified models, it is difficult to capture the complex dynamic changes and nonlinear characteristics in circuit board detection data. This limitation is particularly prominent when dealing with complex circuit board systems, where small abnormal signals may be early signs of major faults, and failure to be discovered and handled in time may lead to serious consequences. Summary of the invention
[0005] The purpose of the present invention is to solve the shortcomings of the prior art and propose a cloud-based circuit board inspection data management system.
[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a cloud-based circuit board detection data management system, the system comprising a signal acquisition module, a dynamic model building module, a rate adjustment prediction module, a nonlinear anomaly recognition module, an anomaly analysis module, an optimization adjustment feedback module, and an abnormal data archiving module; The signal acquisition module uses sensors to collect temperature, current, and voltage parameters of the circuit board, integrates timestamps and device identification information, constructs a data set, and generates a circuit board operation signal data set; The dynamic model building module uses the circuit board operation signal data set to analyze the parameter change trend, describe the interaction between the circuit board performance and environmental variables, and obtain the circuit board dynamic performance interaction model; The rate adjustment prediction module comprehensively analyzes the changes in data flow and processing capacity based on the dynamic performance interaction model of the circuit board, adjusts the data acquisition frequency and processing rate in real time, and generates a comprehensive adjustment plan; The nonlinear anomaly identification module analyzes the circuit board operation signal data set according to the comprehensive adjustment plan, identifies signals that deviate from the normal operation mode, and marks potential abnormal signals to obtain a circuit board abnormal signal mark set; The abnormal analysis module uses the circuit board abnormal signal tag set to analyze each tag, classify the type of abnormal signal, and mine the cause, potential impact and law of the abnormal signal to obtain the abnormal signal analysis result; The anomaly analysis module includes a type classification submodule, a cause mining submodule, and an impact analysis submodule; The type classification submodule traverses each tag based on the abnormal signal tag set of the circuit board, and uses the preset classification logic to determine the type of each abnormal signal. If the characteristics of the abnormal signal match the specified type, it is classified into the associated category. After all abnormal signals are classified, they are sorted and archived to obtain the abnormal signal type classification result; The cause mining submodule is based on the abnormal signal type classification results. By constructing a dynamic model for each abnormal signal, the abnormal signal generation process is regarded as a result driven by internal state and external input. By applying stability analysis and chaos theory, the conditions and laws leading to the abnormality are identified by analyzing the steady state and transition state of time series data, and the conditions and laws are recorded to obtain the abnormal signal cause analysis results. The stability analysis and chaos theory are based on the improved formula I and formula II:
[0007] in, is the rate of change of the system state variables, is the system state variable, To control the input variables, is the external disturbance variable, are the influence coefficients of the system internal state, control input and external disturbance on the system state change rate, which are determined by regression analysis of historical data. is the free term of the model, and the analytical The influence of other factors on the rate of change of system state is determined through data analysis;
[0008] in, are the standard parameters of the Lorenz system, representing the Prandtl number, Rayleigh number and the vertical scale factor of the system, Represent the adjustment values of the system to temperature sensitivity, current responsiveness and voltage stability, respectively, and are determined by genetic algorithm. is the system state variable; represents the derivative with respect to time, and also represents the variables rate of change over time; Identify the conditions and patterns that lead to abnormalities, and record them to obtain the results of abnormal signal cause analysis; The impact analysis submodule evaluates the scope and degree of influence of each abnormal signal on the performance of the circuit board based on the abnormal signal cause analysis results, and predicts the potential impact of the abnormality on the operation of the circuit board through the impact assessment logic with reference to the factors of the strength, frequency and duration of the abnormal signal, and obtains the abnormal signal analysis results; The optimization and adjustment feedback module evaluates and optimizes the current monitoring strategy and parameter settings based on the abnormal signal analysis results, adjusts the signal acquisition accuracy and abnormal identification process, and forms a monitoring strategy optimization and adjustment plan; The abnormal data archiving module organizes and archives key information in the monitoring strategy optimization and adjustment plan and the abnormal signal analysis results, provides a reference for future strategy adjustment and fault prevention, and creates a circuit board abnormal signal analysis information set.
[0009] The present invention has improvements in that the circuit board operation signal data set includes temperature readings, current measurement values, voltage measurement values, data acquisition timestamps, and unique device identifiers; the circuit board dynamic performance interactive model includes parameter change trend graphs, environmental variable impact analysis, and performance impact predictions; the comprehensive adjustment plan includes acquisition frequency adjustment indicators, processing rate adjustment indicators, and data flow change response measures; the circuit board abnormal signal marking set includes abnormal signal type, abnormal occurrence time, abnormal signal strength, and potential impact area; the abnormal signal analysis results include abnormal cause analysis, impact range estimation, and abnormal signal occurrence rules; the monitoring strategy optimization and adjustment plan includes signal acquisition accuracy adjustment strategy, abnormal identification process optimization measures, and monitoring parameter adjustment list; the circuit board abnormal signal analysis information set includes archived abnormal signal information, strategy adjustment records, and fault prevention plans.
[0010] The present invention is improved in that the signal acquisition module includes a sensor deployment submodule, a time synchronization submodule, and a data integration submodule; The sensor deployment submodule uses sensors to collect temperature, current, and voltage parameters of the circuit board, integrates timestamp and device identification information, constructs a data set, and obtains sensor raw data; The time synchronization submodule performs time tag calibration on the sensor raw data, aligns all data points in time, and obtains time synchronization data; The data integration submodule organizes the time synchronization data according to a predetermined format, including merging data with the same time stamp and normalizing data fields, to obtain a circuit board operation signal data set.
[0011] The present invention is improved in that the dynamic model construction module includes a parameter analysis submodule, a model design submodule, and a performance interaction analysis submodule; The parameter analysis submodule performs trend analysis on the temperature, current, and voltage parameters in the circuit board operation signal data set, identifies key parameter changes and periodic fluctuations, and obtains key parameter change trend analysis results; The model design submodule designs a model to describe the relationship between circuit board performance and environmental variables based on the key parameter change trend analysis results, and constructs an initial dynamic performance model; The performance interaction analysis submodule utilizes the initial dynamic performance model to refine and analyze the interaction between circuit board performance and environmental variables, optimizes model parameters and reflects the current interaction, and obtains a circuit board dynamic performance interaction model.
[0012] The present invention is improved in that the rate adjustment prediction module includes a flow monitoring submodule, a capability analysis submodule, and a strategy optimization submodule; The flow monitoring submodule analyzes the fluctuation patterns and trend changes in the data set based on the dynamic performance interaction model of the circuit board, identifies the key change points of data flow and processing capacity, constructs a data change analysis chart, and obtains data change analysis results; The capability analysis submodule introduces the data change analysis results, evaluates whether the current data acquisition frequency and processing rate meet the performance requirements of the circuit board, identifies key change points based on the analysis results, adjusts the baseline values of the acquisition frequency and processing rate, sets threshold adjustment rules to match future changes, constructs an adjustment strategy diagram, and generates an adjustment strategy solution; The strategy optimization submodule adopts an adjustment strategy scheme to plan the real-time adjustment logic of the acquisition frequency and processing rate, including setting the trigger conditions for dynamic adjustment, defining the algorithm for the adjustment range and presetting the optimal adjustment path, compiling the adjustment operation steps, building an execution flowchart, and obtaining a comprehensive adjustment scheme.
[0013] The present invention is improved in that the nonlinear anomaly identification module includes a signal deviation analysis submodule, an anomaly marking submodule, and an anomaly signal set construction submodule; The signal deviation analysis submodule performs a cyclic check on each signal based on the comprehensive adjustment scheme. If the signal data deviates from the set normal operating range, it is recorded as a potential abnormality. The set abnormality judgment logic is used to select and judge the suspected abnormal signal to obtain a deviation signal record set; The abnormal marking submodule analyzes each potential abnormal signal based on the deviation signal record set to determine its deviation degree and frequency. If the deviation pattern meets the characteristics of nonlinear abnormality, it will mark the abnormality, assign a unique identifier to each confirmed abnormal signal, and record the time when the abnormality occurred and the signal strength, and generate an abnormal signal mark list; The abnormal signal set construction submodule organizes and summarizes all marked abnormal signals based on the abnormal signal tag list, including type, occurrence time, signal strength and potential impact area, removes duplicate or erroneous tags, optimizes the structure and format of the tag information, and obtains the circuit board abnormal signal tag set.
[0014] The present invention is improved in that the optimization adjustment feedback module includes a strategy evaluation submodule, a parameter adjustment submodule, and an implementation plan construction submodule; The strategy evaluation submodule evaluates the current monitoring strategy and parameter settings based on the abnormal signal analysis results, determines the efficiency and accuracy of the current strategy in identifying and responding to abnormal signals, identifies the shortcomings of the strategy, and obtains a strategy evaluation improvement point analysis set; The parameter adjustment submodule adjusts the signal acquisition accuracy and abnormality recognition process based on the strategy evaluation improvement point analysis set, and the adjustment includes modifying the acquisition frequency, adjusting the data processing threshold, and optimizing the recognition logic of abnormal signals, constructing parameter adjustment records, and obtaining optimized parameter settings; The implementation plan construction submodule is based on the optimized parameter settings, integrates the current monitoring strategy and the adjusted parameters, verifies its effect and feasibility through simulation tests, and obtains the monitoring strategy optimization and adjustment plan.
[0015] The present invention is improved in that the abnormal data archiving module includes an information sorting submodule, a strategy archiving submodule, and an information set creation submodule; The information sorting submodule collects monitoring strategy optimization and adjustment plans and abnormal signal analysis results, screens and classifies the collected information, removes redundant data, retains key information, and obtains a key information summary; The strategy archiving submodule organizes the contents of the monitoring strategy optimization and adjustment plan and abnormal signal analysis results based on the key information summary, archives them according to chronological order and importance, and generates an archiving strategy document; The information set creation submodule uses the archiving strategy document to perform an archiving operation, stores the sorted key information in a database or a file system according to a predetermined strategy, and creates a circuit board abnormal signal analysis information set.
[0016] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, the nonlinear dynamic system theory is used to significantly improve the system's ability to identify small abnormal signals. By constructing a dynamic system model of circuit board detection data and combining chaos theory with stability analysis of dynamic systems, abnormal patterns can be accurately identified from complex, nonlinearly changing data streams. Not only does it greatly improve the accuracy and sensitivity of fault diagnosis, it also enables the system to detect potential fault signals in a timely manner, thereby greatly reducing the intervention time before the fault develops to a serious stage, effectively preventing possible serious consequences. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 The module diagram of the cloud circuit board detection data management system proposed by the present invention; Figure 2 The system framework diagram of the cloud circuit board detection data management system proposed by the present invention; Figure 3 A schematic diagram of a signal acquisition module in a cloud-based circuit board detection data management system proposed by the present invention; Figure 4 A schematic diagram of a dynamic model building module in a cloud-based circuit board inspection data management system proposed by the present invention; Figure 5 A schematic diagram of a rate adjustment prediction module in a cloud-based circuit board inspection data management system proposed by the present invention; Figure 6 A schematic diagram of a nonlinear anomaly recognition module in a cloud-based circuit board inspection data management system proposed by the present invention; Figure 7 A schematic diagram of an abnormality analysis module in a cloud-based circuit board inspection data management system proposed by the present invention; Figure 8 A schematic diagram of optimizing and adjusting the feedback module in the cloud-based circuit board inspection data management system proposed by the present invention; Fig. 9 The present invention provides a schematic diagram of an abnormal data archiving module in a cloud-based circuit board detection data management system. DETAILED DESCRIPTION
[0018] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0019] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined. Example
[0020] See also Figure 1 , the present invention provides a technical solution: a cloud-based circuit board detection data management system, the system includes a signal acquisition module, a dynamic model building module, a rate adjustment prediction module, a nonlinear anomaly recognition module, an anomaly analysis module, an optimization adjustment feedback module, and an abnormal data archiving module; The signal acquisition module uses sensors to collect the temperature, current, and voltage parameters of the circuit board, integrates the timestamp and device identification information, constructs a data set, and generates a circuit board operation signal data set; The dynamic model building module uses the circuit board operation signal data set to analyze the parameter change trend, describe the interaction between the circuit board performance and environmental variables, and obtain the circuit board dynamic performance interaction model; The rate adjustment prediction module comprehensively analyzes the changes in data flow and processing capacity based on the dynamic performance interaction model of the circuit board, adjusts the data acquisition frequency and processing rate in real time, and generates a comprehensive adjustment plan; The nonlinear anomaly identification module analyzes the circuit board operation signal data set according to the comprehensive adjustment plan, identifies the signals that deviate from the normal operation mode, marks the potential abnormal signals, and obtains the circuit board abnormal signal mark set; The abnormal analysis module uses the abnormal signal tag set of the circuit board to analyze each tag, classify the type of abnormal signal, and explore the cause, potential impact and law of the abnormal signal to obtain the abnormal signal analysis results; The optimization and adjustment feedback module evaluates and optimizes the current monitoring strategy and parameter settings based on the abnormal signal analysis results, adjusts the signal acquisition accuracy and abnormal identification process, and forms a monitoring strategy optimization and adjustment plan; The abnormal data archiving module organizes and archives the key information in the monitoring strategy optimization and adjustment plan and the abnormal signal analysis results, provides a reference for future strategy adjustment and fault prevention, and creates a circuit board abnormal signal analysis information set.
[0021] The circuit board operation signal data set includes temperature readings, current measurement values, voltage measurement values, data acquisition timestamp, and device unique identification. The circuit board dynamic performance interaction model includes parameter change trend charts, environmental variable impact analysis, and performance impact predictions. The comprehensive adjustment plan includes acquisition frequency adjustment indicators, processing rate adjustment indicators, and data flow change response measures. The circuit board abnormal signal marking set includes abnormal signal type, abnormal occurrence time, abnormal signal strength, and potential impact area. The abnormal signal analysis results include abnormal cause analysis, impact range estimation, and abnormal signal occurrence rules. The monitoring strategy optimization and adjustment plan includes signal acquisition accuracy adjustment strategy, abnormal identification process optimization measures, and monitoring parameter adjustment list. The circuit board abnormal signal analysis information set includes archived abnormal signal information, strategy adjustment records, and fault prevention plans.
[0022] See also Figure 2 and Figure 3 ,The signal acquisition module includes a sensor deployment submodule, a time synchronization submodule, and a data integration submodule; The sensor deployment submodule uses sensors to collect the temperature, current, and voltage parameters of the circuit board, integrates the timestamp and device identification information, builds a data set, and obtains the specific process of the sensor raw data as follows: The signal acquisition module is based on the physical characteristics of the circuit board and adopts sensor technology. It uses temperature, current and voltage sensors to collect the temperature, current and voltage parameters of the circuit board respectively, and realizes the preliminary conversion of data through integrated circuits (ICs) and microprocessors (µPs). It combines with timestamp generators to generate accurate event recording time, and uses device identification encoders to assign unique device identification to each data point to generate sensor raw data sets.
[0023] The time synchronization submodule calibrates the time tags of the sensor raw data and aligns all data points in time. The specific process of obtaining time synchronization data is as follows: The time synchronization submodule performs time synchronization based on the Network Time Protocol (NTP). It uses the NTP client to query the NTP server to obtain the standard time. Through the time calibration algorithm, it calibrates the time tag of each data point in the sensor's original data set to ensure the consistency of all data points in time. It uses the time alignment technology to align all data points in time and generate a time-synchronized data set.
[0024] The data integration submodule organizes the time synchronization data according to a predetermined format, including merging data with the same timestamp and normalizing data fields. The specific process of obtaining the circuit board operation signal data set is as follows: The data integration submodule performs data integration based on data fusion technology. It uses a database management system (DBMS) to execute SQL query commands to merge data with the same timestamp, and uses data formatting tools such as Python's Pandas library to normalize data fields, including renaming column names to meet predefined standards and converting data types to ensure consistency, generating a circuit board operation signal data set.
[0025] See also Figure 2 and Figure 4 ,The dynamic model building module includes parameter analysis submodule, model design submodule, and performance interaction analysis submodule; The parameter analysis submodule performs trend analysis on the temperature, current, and voltage parameters in the circuit board operation signal data set, identifies key parameter changes and periodic fluctuations, and obtains the specific process of key parameter change trend analysis results as follows: The parameter analysis submodule is based on the circuit board operation signal data set and adopts the time series analysis method. It uses Python's Pandas and NumPy libraries for data preprocessing, including removing outliers, filling missing values, and standardizing temperature, current, and voltage parameters. Then, it uses moving average and exponential smoothing functions to calculate trend analysis, and defines a 7-day periodic detection through a sliding window to identify periodic fluctuations. It performs statistical analysis on the changing trends of key parameters, and uses the Pearson correlation coefficient to calculate the correlation between parameters to generate key parameter changing trend analysis results.
[0026] The model design submodule is based on the key parameter change trend analysis results, and the design model describes the relationship between the circuit board performance and the environmental variables. The specific process of building the initial dynamic performance model is as follows: The model design submodule adopts machine learning algorithm based on the trend analysis results of key parameter changes, and uses the SciKit-Learn library to select the linear regression model for design. First, the parameters of the linear regression model are defined, including setting the maximum number of iterations to 1000 and the tolerance to 1e-4. The temperature, current, and voltage parameters in the training data set are used as independent variables, and the environmental variables are used as dependent variables. The weights and biases are adjusted through the fitting process to minimize the prediction error. Finally, cross-validation is used to evaluate the accuracy and generalization ability of the model to generate an initial dynamic performance model.
[0027] The performance interaction analysis submodule uses the initial dynamic performance model to refine the analysis of the interaction between the circuit board performance and environmental variables, optimize the model parameters and reflect the current interaction. The specific process of obtaining the circuit board dynamic performance interaction model is as follows: The performance interaction analysis submodule uses the initial dynamic performance model, adopts the system dynamics method, and uses MATLAB's Simulink environment for simulation. First, according to the interaction between the circuit board performance and the environmental variables, the input and output variables of the model are set, including the temperature, current, and voltage parameters of the circuit board as inputs, and performance indicators such as heat dissipation efficiency and power consumption as outputs. Then, the state equation and transfer function of the system are defined. By adjusting the parameters in the Simulink module, such as the gain, integrator, and differentiator settings, the time step is selected as 0.01 seconds to simulate the dynamic response of the circuit board under different environmental conditions. Finally, according to the simulation results, the analysis model parameters are refined and adjusted to better reflect the current performance and environmental variable interaction to generate a circuit board dynamic performance interaction model.
[0028] See also Figure 2 and Figure 5 ,The rate adjustment prediction module includes a traffic monitoring submodule, a capacity analysis submodule, and a ,strategy optimization submodule; The traffic monitoring submodule is based on the dynamic performance interaction model of the circuit board, analyzes the fluctuation pattern and trend changes in the data set, identifies the key change points of data flow and processing capacity, and constructs a data change analysis chart. The specific process of obtaining the data change analysis results is as follows: The flow monitoring submodule is based on the dynamic performance interaction model of the circuit board, adopts the fluctuation analysis algorithm, uses Python's Pandas library for data frame operations, and combines the Matplotlib library to draw data change analysis charts. First, by calculating the moving average and standard deviation, the fluctuation pattern and trend changes in the data set are identified. Then, the Z-score method is used to standardize the data and identify the key change points related to processing power. By setting the Z-score threshold to ±2, the data points that deviate significantly from the average level are identified. Finally, the Matplotlib library is used to draw data change analysis charts to show the key change points of data flow and processing power, and generate data change analysis results.
[0029] The capability analysis submodule introduces the data change analysis results to evaluate whether the current data acquisition frequency and processing rate meet the performance requirements of the circuit board. According to the analysis results, the key change points are identified, the baseline values of the acquisition frequency and processing rate are adjusted, and the threshold adjustment rules are set to match future changes. The adjustment strategy diagram is constructed and the specific process of generating the adjustment strategy plan is as follows: The capability analysis submodule introduces the data change analysis results, adopts a dynamic adjustment algorithm, and uses Python's SciPy library for numerical analysis. First, based on the data change analysis results, it evaluates whether the current data acquisition frequency and processing rate meet the performance requirements of the circuit board, and estimates the optimal acquisition frequency and processing rate under different working conditions through linear interpolation. Then, the least squares method is used to determine the baseline value of the acquisition frequency and processing rate. According to the predicted data traffic change trend, the threshold adjustment rule is set, and the step function is used to adjust the acquisition frequency and processing rate to match future changes. Finally, an adjustment strategy diagram is constructed to generate an adjustment strategy plan.
[0030] The strategy optimization submodule adopts the adjustment strategy scheme to plan the real-time adjustment logic of the acquisition frequency and processing rate, including setting the trigger conditions for dynamic adjustment, defining the algorithm for the adjustment range and presetting the optimal adjustment path, compiling the adjustment operation steps, and building the execution flow chart. The specific process of obtaining the comprehensive adjustment scheme is as follows: The strategy optimization submodule adopts an adjustment strategy scheme, plans real-time adjustment logic, and uses Python's DecisionTree library to build a decision tree model. First, set the trigger conditions for dynamic adjustment based on real-time monitoring data of data flow and processing capacity. When the data point exceeds the preset threshold, the adjustment is triggered. Then, define the algorithm for the adjustment range, use regression analysis based on historical data to determine the adjustment range, preset the optimal adjustment path, and use the decision tree model to select the best adjustment plan based on the current status and prediction results. Finally, compile the adjustment operation steps, use flowchart software such as Visio to draw an execution flowchart, and generate a comprehensive adjustment plan.
[0031] See also Figure 2 and Figure 6 ,The nonlinear anomaly recognition module includes a signal deviation analysis submodule, anomaly marking submodule, and anomaly signal set construction submodule; The signal deviation analysis submodule performs a cyclic check on each signal based on the comprehensive adjustment scheme. If the signal data deviates from the set normal operating range, it is recorded as a potential abnormality. The set abnormal judgment logic is used to select and judge the suspected abnormal signal. The specific process of obtaining the deviation signal record set is as follows: The signal deviation analysis submodule is based on a comprehensive adjustment plan, adopts a statistical anomaly detection algorithm, and uses Python's SciPy library for numerical calculations. First, the statistical threshold of the normal operating range is set, including the mean plus or minus three times the standard deviation, and the temperature, current, and voltage signals of the circuit board are checked cyclically. Then, the Z score of each signal data point is calculated to evaluate the degree of deviation from the mean. If the Z score of the signal data exceeds the preset threshold, it is considered to deviate from the normal operating range, and this data point is recorded as a potential anomaly. Using the anomaly judgment logic, based on the Z score and the signal change trend, the suspected abnormal signals are screened and judged. Finally, the deviation signal record set is sorted, including the signal timestamp, Z score, and degree of deviation, to generate a deviation signal record set.
[0032] The abnormal marking submodule analyzes each potential abnormal signal based on the deviation signal record set to determine its deviation degree and frequency. If the deviation pattern meets the characteristics of nonlinear abnormality, it will be marked as abnormal, and a unique identifier will be assigned to each confirmed abnormal signal. The time when the abnormality occurred and the signal strength are recorded. The specific process of generating the abnormal signal marking list is as follows: The anomaly marking submodule is based on the deviation signal record set, adopts pattern recognition algorithm, and uses Python's Sklearn library to execute the machine learning process. First, the features of each potential abnormal signal in the deviation signal record set are extracted, including the degree of deviation, frequency and duration. Then, based on the characteristics of nonlinear anomalies, such as rapid changes and periodic fluctuations, the support vector machine (SVM) classifier is used to analyze and judge the signal. If the characteristics of the signal conform to the nonlinear abnormal pattern, the signal is marked as abnormal. A unique identifier is assigned to each confirmed abnormal signal, and the time of occurrence, signal strength and duration of the abnormality are recorded. Finally, the abnormal signal marker list is summarized to generate an abnormal signal marker list.
[0033] The abnormal signal set construction submodule organizes and summarizes all marked abnormal signals based on the abnormal signal tag list, including type, occurrence time, signal strength and potential impact area, removes duplicate or erroneous tags, optimizes the structure and format of tag information, and obtains the specific process of the abnormal signal tag set of the circuit board as follows: The abnormal signal set construction submodule is based on the abnormal signal marker list, adopts the data sorting algorithm, and uses Python's Pandas library for data frame operations. First, all the marked abnormal signals are sorted and summarized, including signal type, occurrence time, signal strength and potential impact area. Then, the abnormal signal marker list is checked to remove duplicate or erroneous markers. Through data cleaning methods such as deleting duplicates and filling missing values, the structure and format of the marker information are optimized. Finally, a structured circuit board abnormal signal marker set is constructed to facilitate subsequent analysis and processing, and a circuit board abnormal signal marker set is generated.
[0034] See also Figure 2 and Figure 7 ,The anomaly analysis module includes type classification submodule, cause mining submodule, and impact analysis submodule; The type classification submodule traverses each tag based on the abnormal signal tag set of the circuit board, and uses the preset classification logic to determine the type of each abnormal signal. If the characteristics of the abnormal signal match the specified type, it is classified into the associated category. After all abnormal signals are classified, they are sorted and archived. The specific process of obtaining the abnormal signal type classification result is as follows: The type classification submodule is based on the abnormal signal tag set of the circuit board, adopts the decision tree algorithm, and uses the scikit-learn library in the Python language to implement the decision tree classifier. The decision tree parameters are defined, including the criterion set to gini to measure the split quality, and the max_depth limit to the maximum depth of the tree to avoid overfitting. Each tag is traversed, and the fit method is executed to train the model according to the feature vector of the abnormal signal, including signal strength, frequency, and duration. The predict method is used to judge the type of the new abnormal signal. If the feature vector of the abnormal signal matches the decision path of a specific type, it is classified into the associated category. After all abnormal signals are classified, the sort method is executed to sort and archive the abnormal signal types to generate the abnormal signal type classification results; The cause mining submodule is based on the classification results of abnormal signal types. By constructing a dynamic model for each abnormal signal, the abnormal signal generation process is regarded as the result driven by internal state and external input. By applying stability analysis and chaos theory, the conditions and laws that lead to abnormalities are identified by analyzing the steady state and transition state of time series data, and the conditions and laws are recorded. The specific process of obtaining the cause analysis results of abnormal signals is as follows: The cause mining submodule is based on the abnormal signal type classification results, adopts dynamic system theory, uses Python's NumPy library for numerical calculations, builds a dynamic system model for each abnormal signal, defines model parameters, including model state variables, control input variables, and external disturbance variables, and executes the odeint function to simulate the system dynamic behavior. During the simulation process, stability analysis and chaos theory are applied to perform Lorenz system analysis on the simulated time series data, and the matplotlib library is used to draw a trajectory diagram of the system state changing over time. Through trajectory analysis, the conditions and laws that lead to the anomaly are identified and recorded to generate the abnormal signal cause analysis results; Stability analysis and chaos theory are based on the improved formula I and formula II:
[0035] in, Is the rate of change of the system state variables; it indicates how fast the system state changes over time. It is a system state variable; it represents the current state of the circuit board, such as temperature, current or voltage. Control input variable; refers to the control signal or action applied to the circuit board, such as a change in the power supply. It is an external disturbance variable; it reflects the impact of external environmental factors on the state of the circuit board, such as changes in ambient temperature. They are the influence coefficients of the system internal state, control input and external disturbance on the system state change rate, which are determined through regression analysis of historical data; ensuring that the model accurately reflects the actual system behavior. is the free term of the model, and the analytical The influence of other factors on the rate of change of system state is determined through data analysis;
[0036] in, are the standard parameters of the Lorenz system, representing the Prandtl number, Rayleigh number and the vertical scale factor of the system respectively; the parameters are used in the original Lorenz system to describe the basic characteristics of fluid convection, heat conduction and spatial structure. They represent the adjustment values of the system's temperature sensitivity, current responsiveness, and voltage stability, respectively, and are determined through a genetic algorithm; making the model closer to the actual dynamic behavior of the circuit board.
[0037] It is the system state variable; it is also interpreted as different state indicators or characteristics of the circuit board. The dynamic behavior of the circuit board and the dynamic process of abnormal signal generation are simulated through a set of equations. represents the derivative with respect to time, and also represents the variables Rates of change over time. These rates of change are key components in dynamic models and describe how quickly the state of a system evolves over time. In fluid convection or similar physical processes, this concept of time derivatives is used to capture instantaneous changes in the state of a system, making it possible to simulate the dynamic behavior of the system over time.
[0038] Identify the conditions and patterns that lead to anomalies, record them, and obtain the results of the abnormal signal cause analysis.
[0039] The execution process is as follows: By analyzing the historical operation data of the circuit board, the coefficients in the dynamic system equations are determined using methods such as linear regression. , ensuring that the model accurately reflects the actual response of the board under different control inputs and external conditions.
[0040] Using the determined parameters, dynamic system equations and Lorenz system equations are simulated, and the changes of the system over time under given initial conditions are calculated through numerical solution methods such as odeint.
[0041] Tuning parameters of Lorenz system using genetic algorithm Optimization is performed to ensure that the model simulation results are closest to the actual observed abnormal behavior of the board.
[0042] Through simulation and parameter optimization, the obtained model can reveal the causes of abnormal signals in the circuit board and provide a scientific basis for further fault diagnosis and prevention.
[0043] Through the above process, not only the accuracy of cause analysis is improved, but also the model's adaptability to the unique dynamic characteristics of circuit boards is enhanced, providing a more accurate and scientific method for the identification and processing of abnormal signals of circuit boards.
[0044] The impact analysis submodule evaluates the scope and degree of impact of each abnormal signal on the performance of the circuit board based on the abnormal signal cause analysis results. It refers to the factors of the strength, frequency and duration of the abnormal signal, and predicts the potential impact of the abnormality on the operation of the circuit board through the impact assessment logic. The specific process of obtaining the abnormal signal analysis results is as follows: The impact analysis submodule adopts the influence analysis method based on the abnormal signal cause analysis results, uses Python's Pandas library for data processing, defines the impact assessment model, includes abnormal signal strength, frequency, duration and other influencing factors as input parameters, and executes the DataFrame.apply method to quantitatively evaluate the impact range and degree of each abnormal signal. During the quantification process, a custom function is used to calculate the possible impact of the abnormal signal on the performance of the circuit board according to the different types and causes of the abnormal signal. The sort_values method is executed to sort all abnormal signals by the degree of impact to generate the abnormal signal analysis results.
[0045] See also Figure 2 and Figure 8 ,The optimization adjustment feedback module includes a strategy evaluation submodule, a parameter adjustment submodule, and an implementation plan building submodule; The strategy evaluation submodule evaluates the current monitoring strategy and parameter settings based on the abnormal signal analysis results, determines the efficiency and accuracy of the current strategy in identifying and responding to abnormal signals, identifies the shortcomings of the strategy, and obtains the strategy evaluation improvement point analysis set. The specific process is as follows: The strategy evaluation submodule uses the decision analysis algorithm based on the abnormal signal analysis results and uses Python's DecisionTreeClassifier library to build a decision tree model. First, the parameters of the decision tree model are defined, including a maximum depth of 5 layers and information gain selected as entropy. The data in the abnormal signal analysis results, including the abnormal type, frequency, duration and impact range, are used as input features, and the efficiency and accuracy of the current monitoring strategy and parameter settings are used as target variables. Then, the decision tree model is trained, the decision path output by the model is analyzed, and the shortcomings of the identification strategy are identified, such as nodes with slow identification speed or low accuracy. Finally, the identified strategy improvement points are sorted out, including increasing the frequency of data collection, adjusting the abnormal identification threshold and optimizing the data processing process, to generate a strategy evaluation improvement point analysis set.
[0046] The parameter adjustment submodule adjusts the signal acquisition accuracy and abnormality recognition process based on the strategy evaluation improvement point analysis set. The adjustment includes modifying the acquisition frequency, adjusting the data processing threshold, optimizing the recognition logic of abnormal signals, and constructing parameter adjustment records. The specific process of obtaining the optimized parameter settings is as follows: The parameter adjustment submodule is based on the strategy evaluation improvement point analysis set and adopts a parameter optimization algorithm. It uses the optimize module in Python's SciPy library to adjust parameters. First, the objective function of parameter adjustment is set to minimize the response time of anomaly recognition and maximize the accuracy. The deficiencies identified in the strategy evaluation improvement point analysis set, including the acquisition frequency and data processing threshold, are used as adjustment variables. Then, the sequential quadratic programming (SQP) method is used to find the optimal solution, and the acquisition frequency, data processing threshold and abnormal signal recognition logic are dynamically adjusted. Finally, the results of each parameter adjustment are recorded, including the changes in the acquisition frequency, processing threshold and recognition logic before and after the adjustment, to generate the optimized parameter settings.
[0047] The implementation plan building submodule integrates the current monitoring strategy and adjusted parameters based on the optimized parameter settings, verifies its effectiveness and feasibility through simulation tests, and obtains the specific process of monitoring strategy optimization and adjustment plan as follows: The implementation plan builds a sub-module based on the optimized parameter settings, adopts simulation test verification algorithm, and uses MATLAB's Simulink environment for system-level simulation. First, according to the optimized parameter settings, including the adjusted data acquisition frequency, processing rate and abnormal signal recognition logic, the corresponding parameters of the Simulink model are configured. Then, the input conditions of the simulation test are defined, including normal operation signals and artificially generated abnormal signals, to verify the effect and feasibility of the monitoring strategy and parameter adjustment. Through simulation operation, the system response time, recognition accuracy and system stability data are collected. Finally, according to the simulation test results, the current monitoring strategy and adjusted parameters are integrated to evaluate their performance in actual applications and generate monitoring strategy optimization and adjustment plans.
[0048] See also Figure 2 and Fig. 9 ,The abnormal data archiving module includes an information ,collation submodule, a policy archiving submodule, and an information set creation submodule; The information collation submodule collects monitoring strategy optimization and adjustment plans and abnormal signal analysis results, screens and classifies the collected information, removes redundant data, retains key information, and obtains the specific process of key information summary as follows: The information collation submodule collects monitoring strategy optimization and adjustment plans and abnormal signal analysis results, adopts data screening algorithms, and uses Python's Pandas library for data frame operations. First, the data set of monitoring strategy optimization and adjustment plans and abnormal signal analysis results is read. Then, Boolean indexing and conditional filtering methods are applied to remove redundant items and irrelevant information in the data, such as duplicate strategy records and irrelevant analysis data points. Subsequently, the remaining data is classified and summarized, and key information summaries are sorted out according to the time sequence and the importance of the information, including improved strategy points, key abnormal signal types and impact ranges, to generate key information summaries.
[0049] The strategy archiving submodule organizes the monitoring strategy optimization and adjustment plan and abnormal signal analysis results based on the key information summary, and archives them in chronological order and importance. The specific process of generating the archiving strategy document is as follows: The strategy archiving submodule is based on the key information summary and adopts the document management algorithm. It uses Python's Docx library to create and edit Word documents. First, according to the content of the key information summary, the structure and format of the document are designed, including the document title, chapter title and content list. Then, the key information summary is converted into document content, arranged and compiled according to chronological order and information importance, such as listing the optimization points of each monitoring strategy and the corresponding abnormal signal analysis results according to the timeline. Subsequently, the document is reviewed and formatted to ensure the accuracy and readability of the information. Finally, the document is saved in Word format to generate an archiving strategy document.
[0050] The information set creation submodule uses the archiving strategy document to perform the archiving operation and stores the sorted key information in the database or file system according to the established strategy. The specific process of creating the circuit board abnormal signal analysis information set is as follows: The information set creation submodule utilizes the archiving strategy document, adopts the database storage algorithm, and uses the MySQL database for data storage operations. First, the table structure of the database is designed, including the table name, field name, and data type, to adapt to the information structure in the archiving strategy document, such as the monitoring strategy optimization point table and the abnormal signal analysis result table. Next, SQL statements are written to convert the key information in the archiving strategy document into database records, including insert, update, and query operations. Subsequently, SQL statements are executed to store the document content in the database to ensure data persistence and queryability. Finally, a data backup and recovery strategy is created to prevent data loss or damage, and a circuit board abnormal signal analysis information set is generated.
[0051] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.
Claims
1. Cloud-based circuit board inspection data management system, characterized by: The system comprises: The signal acquisition module uses sensors to collect circuit board parameters, integrates timestamp and device identification information, and generates a circuit board operation signal data set; The dynamic model building module uses the circuit board operation signal data set to analyze parameter change trends, describe the interaction between circuit board performance and environmental variables, and obtain a circuit board dynamic performance interaction model; The rate adjustment prediction module comprehensively analyzes the changes in data flow and processing capacity based on the dynamic performance interaction model of the circuit board, adjusts the data acquisition frequency and processing rate in real time, and generates a comprehensive adjustment plan; The nonlinear anomaly identification module analyzes the circuit board operation signal data set according to the comprehensive adjustment scheme, marks potential abnormal signals, and obtains a circuit board abnormal signal mark set; The abnormality analysis module uses the abnormal signal tag set of the circuit board to explore the causes, potential impacts and rules of the abnormal signal and obtain the abnormal signal analysis results; The optimization and adjustment feedback module evaluates and optimizes the current monitoring strategy and parameter settings based on the abnormal signal analysis results, adjusts the signal acquisition accuracy and abnormality identification process, and forms a monitoring strategy optimization and adjustment plan; The abnormal data archiving module organizes and archives the key information in the monitoring strategy optimization and adjustment plan and the abnormal signal analysis results, and creates a circuit board abnormal signal analysis information set.
2. The cloud-based circuit board inspection data management system according to claim 1 is characterized in that: The circuit board operation signal data set includes temperature readings, current measurement values, voltage measurement values, data acquisition timestamps, and unique device identifiers. The circuit board dynamic performance interaction model includes parameter change trend graphs, environmental variable impact analysis, and performance impact predictions. The comprehensive adjustment plan includes acquisition frequency adjustment indicators, processing rate adjustment indicators, and data flow change response measures. The circuit board abnormal signal marking set includes abnormal signal type, abnormal occurrence time, abnormal signal strength, and potential impact area. The abnormal signal analysis results include abnormal cause analysis, impact range estimation, and abnormal signal occurrence rules. The monitoring strategy optimization and adjustment plan includes signal acquisition accuracy adjustment strategy, abnormal identification process optimization measures, and monitoring parameter adjustment list. The circuit board abnormal signal analysis information set includes archived abnormal signal information, strategy adjustment records, and fault prevention plans.
3. The cloud-based circuit board inspection data management system according to claim 1 is characterized in that: The signal acquisition module comprises: The sensor deployment submodule uses sensors to collect the temperature, current, and voltage parameters of the circuit board, integrates the timestamp and device identification information, builds a data set, and obtains the sensor raw data; The time synchronization submodule performs time tag calibration on the raw data of the sensor, aligns all data points in time, and obtains time synchronization data; The data integration submodule organizes the time synchronization data according to a predetermined format, including merging data with the same timestamp and normalizing data fields, to obtain a circuit board operation signal data set.
4. The cloud-based circuit board inspection data management system according to claim 1, characterized in that: The dynamic model building module includes: The parameter analysis submodule performs trend analysis on the temperature, current, and voltage parameters in the circuit board operation signal data set, identifies key parameter changes and periodic fluctuations, and obtains key parameter change trend analysis results; The model design submodule designs a model to describe the relationship between the circuit board performance and the environmental variables based on the key parameter change trend analysis results, and constructs an initial dynamic performance model; The performance interaction analysis submodule utilizes the initial dynamic performance model to refine and analyze the interaction between circuit board performance and environmental variables, optimize model parameters and reflect current interactions, and obtain a circuit board dynamic performance interaction model.
5. The cloud-based circuit board inspection data management system according to claim 1, characterized in that: The rate adjustment prediction module includes: The flow monitoring submodule analyzes the fluctuation patterns and trend changes in the data set based on the dynamic performance interaction model of the circuit board, identifies the key change points of data flow and processing capacity, constructs a data change analysis chart, and obtains data change analysis results; The capability analysis submodule introduces the data change analysis results, evaluates whether the current data acquisition frequency and processing rate meet the performance requirements of the circuit board, identifies key change points based on the analysis results, adjusts the baseline values of the acquisition frequency and processing rate, sets threshold adjustment rules to match future changes, constructs an adjustment strategy diagram, and generates an adjustment strategy solution; The strategy optimization submodule adopts the adjustment strategy scheme to plan the real-time adjustment logic of the acquisition frequency and processing rate, including setting the trigger conditions for dynamic adjustment, defining the algorithm for the adjustment range and presetting the optimal adjustment path, compiling the adjustment operation steps, building an execution flowchart, and obtaining a comprehensive adjustment plan.
6. The cloud-based circuit board inspection data management system according to claim 1, characterized in that: The nonlinear anomaly recognition module comprises: The signal deviation analysis submodule performs a cyclic check on each signal based on the comprehensive adjustment scheme. If the signal data deviates from the set normal operating range, it is recorded as a potential abnormality. The set abnormality judgment logic is used to select and judge the suspected abnormal signal to obtain a deviation signal record set. The abnormal marking submodule analyzes each potential abnormal signal based on the deviation signal record set to determine its deviation degree and frequency. If the deviation pattern meets the characteristics of nonlinear abnormality, it will mark the abnormality, assign a unique identifier to each confirmed abnormal signal, and record the time when the abnormality occurs and the signal strength, and generate an abnormal signal mark list; The abnormal signal set construction submodule organizes and summarizes all marked abnormal signals based on the abnormal signal mark list, including type, occurrence time, signal strength and potential impact area, removes duplicate or erroneous marks, optimizes the structure and format of the marking information, and obtains the circuit board abnormal signal mark set.
7. The cloud-based circuit board inspection data management system according to claim 1, characterized in that: The abnormality analysis module comprises: The type classification submodule traverses each tag based on the abnormal signal tag set of the circuit board, and uses the preset classification logic to determine the type of each abnormal signal. If the characteristics of the abnormal signal match the specified type, it is classified into the associated category. After all abnormal signals are classified, they are sorted and archived to obtain the abnormal signal type classification results; Based on the abnormal signal type classification results, the cause mining submodule constructs a dynamic model for each abnormal signal, regards the abnormal signal generation process as a result driven by internal state and external input, applies stability analysis and chaos theory, analyzes the steady state and transition state of time series data, identifies the conditions and laws that lead to the anomaly, and records the conditions and laws to obtain the abnormal signal cause analysis results; The impact analysis submodule evaluates the scope and degree of influence of each abnormal signal on the performance of the circuit board based on the abnormal signal cause analysis results, and predicts the potential impact of the abnormality on the operation of the circuit board through the impact assessment logic with reference to the factors of the strength, frequency and duration of the abnormal signal to obtain the abnormal signal analysis results.
8. The cloud-based circuit board inspection data management system according to claim 7, characterized in that: The stability analysis and chaos theory are based on the improved formula I and formula II:
9. Among them, The rate of change of the system state variables, is the system state variable, To control the input variables, is the external disturbance variable, are the influence coefficients of the system internal state, control input and external disturbance on the system state change rate, which are determined by regression analysis of historical data. is the free term of the model, and the analytical The influence of other factors on the rate of change of system state is determined through data analysis; 10. Among them, are the standard parameters of the Lorenz system, representing the Prandtl number, Rayleigh number and the vertical scale factor of the system, Represent the adjustment values of the system to temperature sensitivity, current responsiveness and voltage stability, respectively, and are determined by genetic algorithm. is the system state variable; represents the derivative with respect to time, and also represents the variables rate of change over time; Identify the conditions and patterns that lead to anomalies, record them, and obtain the results of the abnormal signal cause analysis.
Citation Information
Cited By
Gas meter circuit board detection method, system, equipment and medium
CN119375684A
Circuit board sensor nondestructive testing method based on big data
CN119397468A
Street lamp state remote monitoring system and method based on big data
CN119807895A
A remote monitoring system and method for street lamp status based on big data
CN119807895B
Method and system for identifying abnormal region of power transmission line based on infrared thermal image feature fusion
CN119810401A
